LMP90100MHE National Semiconductor, LMP90100MHE Datasheet - Page 27

• Manufacturer Part Number: LMP90100MHE
• Description: AFE, SENSOR, 28TSSOP
• Manufacturer: National Semiconductor

Specifications of LMP90100MHE

• Brief Features: 24bit Low Power Sigma Delta ADC, Automatic Channel Sequencer
• Supply Voltage Range: 2.7V To 5.5V
• Operating Temperature Range: -40°C To +105°C
• Digital Ic Case Style: TSSOP
• No. Of Pins: 28
• Base: RoHS Compliant
• Ic Function: Multi-Channel, Low Power 24-bit Sensor AFE
• Rohs Compliant: Yes

conditions in the system and by appropriate programming of
the System Calibration Control Register (SCALCN bits).
The system zero-scale calibration must be performed prior to
the full-scale calibration and both need to be repeated when
the gain (or the signal path) is changed.
The System Gain coefficient can be filled in even if system
full-scale calibration setup cannot done but instead, a setup
can be done where a known fraction of the full scale (like 0.75
times full scale or 1.25 times full scale) is applied. Here, the
fractional information is entered in the System Gain coeffi-
cient register (in 1.23 fixed-point format) before the calibration
is initiated. The device will then automatically compute the
System Gain coefficient and overwrite the register with the
computed value. This way, one can make use of other known
reference inputs, even if they are not full scale inputs.
The computed calibration coefficients are accurate only to the
effective resolution of the device and will probably contain
some noise. The noise factor can be minimized by computing
over many times, averaging (externally) and putting the re-
sultant values back into the registers.
There are four distinct sets of System Calibration Offset and
Gain Coefficient Registers for use with Channels 0-3. Chan-
nels 4−6 will reuse the System Calibration Offset and Gain
Coefficient values of channels 0−2 respectively.
There are three system calibration coefficients: offset, gain,
and scale. A data-flow diagram of these coefficients and their
register names can be seen in
Types of System Calibration
As seen in
tion: automatic and manual. The automatic calibration occurs
when the LMP90100 computes the offset or gain coefficient
once, stores it in the appropriate registers, and continuously
calibrates the system using this coefficient.
The manual calibration occurs when the user has to manually
compute the offset coefficient or expected output code once,
and store it in the appropriate registers. After this process,
LMP90100 will use this coefficient to continuously calibrate
for the system errors.
Automatic Offset Calibration
Follow the steps below to automatically compute the system
offset coefficient:
1.
2.
FIGURE 13. System Calibration Data-Flow Diagram
Apply a zero scale condition for CH0, CH1, CH2, or CH3
(CHx_INPUTCN).
If the device is to be used in one of the four high ODRs
(214.65, 107.325, 53.6625, or 26.83125 SPS) run the
device at 26.83125 SPS. If the device is to be used in
one of the four low ODRs (13.42, 6.71, 3.355, 1.6775)
run the device at 1.6775 SPS to reduce noise
(CHx_CONFIG: ODR_SEL bit)
Figure
11, there are two types of system calibra-
Figure
13.
30139531
27
3.
4.
5.
6.
Automatic Gain Calibration
Follow the steps below to automatically compute the system
gain coefficient:
1.
2.
3.
4.
5.
6.
Manual Offset Calibration
The “Automatic Offset Calibration” data is accurate only to the
effective resolution of the device and will probably contain
some variation or noise. This probable variation can be re-
duced by performing the “Manual Offset Calibration”.
In this mode, the coefficient has to be known and entered into
the appropriate registers. One way in which the user can
manually compute the coefficient is shown in the following
steps:
1.
2.
3.
4.
The computed offset coefficient has to be a 24-bit two‘s-com-
plement number. For example, if the offset is 200µV for VREF
= 4.1V, then the offset coefficient = [(200µV)(2
= 409d = 0x199.
If the offset is -200µV for VREF = 4.1V, then the offset coef-
ficient is [(-200µV)(2
Manual Gain Calibration
Another way to use the system gain calibration is program-
ming the expected output code and letting LMP90100 cali-
brate itself. For example, suppose VIN =¾(VREF), and thus
the expected output code is 0x60_0000.
1.
2.
3.
Enter the “System Calibration Offset Coefficient
Determination” mode (SCALCN bits)
LMP90100 starts a fresh conversion.
LMP90100 computes the system offset calibration
coefficient once, stores this coefficient in the
CHx_SCAL_OFFSET registers, and continuously
calibrates the system using this coefficient.
LMP90100 exits the “System Calibration Offset
Coefficient Determination” mode.
Repeat the "Automatic Offset Calibration" mode to
calibrate for the offset error first
Apply a full scale condition in which VIN = VINP - VINN
= VREF for CH0, CH1, CH2, or CH3 (CHx_INPUTCN)
Enter the “System Calibration Gain Coefficient
Determination” mode (SCALCN bits)
LMP90100 starts a fresh conversion.
LMP90100 computes the system gain calibration
coefficient once, stores this coefficient in the
CHx_SCAL_GAIN registers, and continuously calibrates
the system using this coefficient.
LMP90100 exits the “System Calibration Gain Coefficient
Determination” mode.
Repeat the "Automatic Offset Calibration" mode and
externally store this coefficient
Repeat step 1 until a finite amount of coefficients are
obtained
Compute the average offset coefficient and enter this
value in the CHx_SCAL_OFFSET registers
LMP90100 will use this coefficient to continuously
calibrate for the offset error.
Apply VIN = VINP - VINN = ¾ (VREF) for CH0, CH1,
CH2, or CH3 (CHx_INPUTCN)
Program the expected output code 0x60_0000 in the
CHx_SCAL_GAIN registers
Enter the “System Calibration Gain Coefficient
Determination” mode (SCALCN bits)
24
)] / [(2)(4.1V)] = -409d = 0XFF_FE67.
24
)] / [(2)(4.1V)]
www.national.com